Since the early days, hurricanes have been a source of mystery, and frontiers of discovery still exist with these powerful storms, especially with the uniqueness of those that find their way to New England. Modern tools make the job of witnessing a storm so much easier, so there will probably never be another storm that 'sneaks up' on the Southern New England coastline. But there will still yet be surprising uncertainty in forecast paths, intensities, and timings, which can make all the difference through the irregular coastline and diverse inland terrain.

The Past of Hurricane Forecasting

"Our New England Hurricanes are a unique breed. I can't get that across enough," Cautions David Vallee, Hydrologist-in-Charge of the New England River Forecast Center, based in Taunton, MA. "We're not going to have 5 days to prepare. Once that acceleration kicks in...we're going to have less than 24-hours." The earliest reports of hurricanes affecting colonists date back to the early 1600s. Hurricanes there were only experienced, and quite misunderstood. It wasn't until Benjamin Franklin, in 1743, when disgusted that he couldn't see a lunar eclipse due to a hurricane outside, deduced that hurricanes didn't actually sit over one spot (he heard that the eclipse was visible in Boston...until a hurricane struck later that night...and he put two-and-two together). In 1821, W.C. Redfield figured out that hurricanes don't travel according to their wind directions. It was also around this time that ships traffic across the Atlantic became frequent enough so as to string together their log books to re-create the path of a storm long since faded. There still wasn't any real way to actually track a hurricane when the Weather Bureau came into being in the 1870s (unless, of course, it was already affecting land).

Not until the early 1900s, with the advent and implementation of the teletype wire can transmit communication from a ship to the shoreline, could a weather bureau get any sort of near real-time weather information on a storm (though it would be not much more than pressure data and a sense of the winds-- if the ship was stable enough to think about sending such a report). Likewise, if there happened to be no ships sailing (voluntarily or involuntarily) through the path of a hurricane, its position, strength, etc, would be completely lost until some new unfortunate encounter. Nevertheless, by this point, forecasters were getting some hints of patterns that existed with hurricane routes and strengthening, which helped greatly in having a chance at forewarning citizens rather than mere observation like an animal in the wild. In the 1930s, the weather balloon, or 'radiosonde', became an instrument of choice, because until that point no one had any idea of the nature of the upper air weather conditions, which just so happen to be critical. The Great 1938 Hurricane was still able to strike with surprise however, because this network was still confined to land...and when you harbor all ships in port because of a hurricane, there goes your information source.

The 1938 Hurricane was a turning point toward modern forecasting. It became immediately apparent that current data from as many sources, locations, and altitudes was of profound importance. Some of the first aircraft reconnaissance flights took place in the 1944 hurricane that struck New England. With all the new technology wrapped up in the 'space race' of the 1960s, much of it trickled down into weather forecasting. Hurricane Donna, in 1960, was one of the first storms seen on Satellite (and radar coincidentally). Today, the tools for peering into the heart of a storm anywhere in the world seem so limitless that it's rather amazing to consider how we ever got along without them-- not too long ago.

New Developments in Tropical Meteorology

The Great Hurricane of 1938, and the corresponding exploration into the character of such a devastating storm, was only the beginning of the finer things of tropical meteorology as it relates to New England storms. More now is known about the conditions required to steer a hurricane from the Western Atlantic, and about the time it takes to get it here. "A hurricane that makes it to the Bahamas...is our business." says Bob Thompson, Meteorologist-in-Charge at the National Weather Service Office in Taunton, MA. Though such a distance between storm and land could be a week's worth of time, it's become all too clear that there is far less lead time with New England storms.

"Hurricanes that make their way toward New England are undergoing transition," says David Vallee. This transition takes a tropical storm and changes it into something more akin to the Blizzards we experience in the Winter (well...without the snow). It's only been in the past few decades that serious research could delve into this with provable results beyond well-reasoned conjecture. Hurricanes, when migrating out of the tropics, make this transition in a sort-of 'know-it-when-you-see-it' kind of way, beginning when they start to interact with non-optimal conditions that are no longer able to stabilize or strengthen a tropical system (cooler water temperatures, jet stream interaction, vertical shear, etc.) Sometimes, they just begin to weaken toward dissipation altogether, especially when caught in an environment that lends no life-line to a hurricane. However, during the peak of the hurricane season (September/October), enough surface heating has taken place over the Summer to enable the region of tropical sustenance in the open waters to be extended northward. Likewise, the approaching Fall season presents an invigorated jet stream pattern from the north, one that encourages large-scale storm formation of a non-tropical nature (called 'cyclogenesis') farther south. It is most often possible in these peak-season months, for a storm to rage as a hurricane, enter waters no longer favorable for intensification (of a tropical nature), but then for it to completely 'transition' itself to a new source of energy (cyclogenesis) and hardly lose outward strength at all during the process. Understanding an approaching hurricane, and what its internal character is, or is trying to become, is essential to figuring out what might happen next.

Most of the major hurricanes that make land-fall in New England will be of this transitioning type. Losing some tropical characteristics, not yet a cyclone, but able to sustain itself by drawing on a hybridized and expanded pool of energy sources. These hurricanes do have a pattern to them. "Our hurricanes have coastal fronts," says David Vallee. "Think coastal fronts...think Winter storm." Indeed, 'Noreasters and hurricanes have exhibited similar qualities. Both bring copious amounts of moisture onshore, and focus it to the west (left) of their track; both generate high winds and gusts, surges and coastal flooding-- all to the eastern (right) side of their track. Normal, tropical hurricanes are symmetric, this sort of thing is not normally seen over some island in the Caribbean.

New Developments in Hydrology

Being in Western New England, understanding the hydrology of an approaching hurricane is even more critical. Our typical left-of-track positioning puts us right in line for flooding threats. As the wind flow moves around the storm, it brings moisture onshore, and lifts it up the mountains, enhancing rainfall even more.

If and when flooding rains linger over an area for a long period of time, all the rivers in a watershed area would fill to capacity and flood. Eventually those swollen streams would combine into the main river of the watershed (ie, the Connecticut River) and drive it well out of its banks for days on end. Forecasters would first have to watch for flash-flooding (quick, localized, intense flooding that strikes and recedes down-river in mere hours), then would monitor and assess all the rainfall across a basin and calculate how high the river will get (called a 'crest'), They would issue warnings and advisories to the communities along those rivers as the crest of water migrates downstream (indeed throughout the time a river is above flood stage)...and then would do the same for the major river in the watershed. If all the water from a particular storm falls on a different side of a particular mountain range, a difference of mere miles, then a completely different river (and a new set of communities are affected). In a particularly large storm situation, multiple basins and watersheds could experience record flooding.

Meteorologists have been slowly acquiring more and more data points in more remote places to have a real-time view of how the rivers and streams are filling up. "We need that base data," Vallee says. "We also use weather radars extensively." The addition of flood control dams since the great floods from 1936 and 1938 is another piece to the puzzle. "They're designed to handle a 6"-8" run-off event," Vallee explains. "They involve the storage of water in flood events, and the graceful release of water after...They're not designed to prevent flooding, but rather reduce the magnitude. It takes 3-5 days for fallen precipitation to make its way down to and through the Connecticut River. The Army Corps of Engineers (who run the flood control dams) and the River Forecast Center are always in close contact, and they are keying in their decisions off of our flood potential forecasts."

New Developments in Technology

"Will we see it coming...absolutely, unequivocally yes...Will we know it will be a major event...yes." Vallee declares. "Our forecasters here have tools...we've trained them (in) what to look for." The biggest improvement to actual forecasting has come with the development of modeling/simulation of the weather, beginning with the roll-out of the first super-computers (about the only things that can handle such a multitude of complex equations). Modeling of future weather conditions is a helpful visual tool to understanding how the atmosphere is interacting with a hurricane, and how rainfall run-off is interacting with the terrain. Only recently has computer modeling taken another step with the ability to run several models of the same kind simultaneously, varying the initial conditions slightly each time, to get a diversified picture of all the options, and therefore the most likely ones. This ensemble modeling technique can be further improved by either improvements to the mathematics that drive it, the computers that crunch them out, the terrain resolution their solving for, or simply with more information-- and all of that is being worked on. Vallee adds, "There are several rivers we're looking to add into our models...for example the Deerfield River...Next year, the Mill River...More rivers (into the models) is always better for our forecasting."

Uncertainty Remains

"I would give anything, to see one satellite picture of what the 1938 hurricane looked like, just so I know what it'll look like again." Dave Vallee says, admitting his humbled feelings openly. "We do so much in our field by touch and feel, and by our experience...and when it comes to a big hurricane in the Northeast...it's very very limited. You have a generation that are the boat-owners...that are the home-owners...that are going to be impacted...We have a very warped reality...a very limited experience."

The trick with hurricanes now is not so much the actual prediction of a strike, but now the location and intensity. In New England, the coastline has many juts, channels, bays, and inlets, many themselves formed after hurricane impacts, and that complicates the arrival of a storm immensely. A slight jog of the track to one side or the other, means someone else is getting life-threatening storm surge. "One of the biggest challenges my office is going to be faced with...is accurately predicting where that axis of heavy rain is going to fall." Vallee says. "If we are one watershed off...one river basin off...it's a bad forecast. That's the state of the science. We know that computer models are not going to forecast this correctly every time. To this day we still do not do a really good job of modeling the intricacies within the clouds...the very producers of this heavy rain. We're being asked to (make these critical forecasts) two days in advance, with a hurricane moving at 60 miles per hour-- The ultimate challenge."

"Maybe we'll over-forecast just a little bit," Vallee concludes. "But I'd rather be in the position of having too many rivers forecasted for flooding, than to have no one receive warning, and we get devastated."